HIDDEN, FOURTH AND SMALLER GLACIERS 149 existed on the floor of the fosse, thus proving that Hidden Glacier, apparently ending in a clean ice slope at an angle of 15° or 20°, in reality extended out beneath the fosse and under the gravel plain. The presence of ice even farther out was plainly shown by the pitted condition of the outwash gravel plain surface, as was observed by Gilbertx in 1899 before the fosse had developed; and it was exhibited with remarkable perfection in 1905. Some of the kettles were small, shallow pits a few feet in diameter. Others were of large size, the largest measuring 250 feet in length and 50 feet in width. The depth of the kettles varied from 1 or 2 feet to 15 or 18 feet. Many of the kettles, especially the shallow ones, were dry; but others, notably the large kettles, contained pools (PI. LXXVI) of clear, deep-blue water a few feet in depth, the greatest depth measured being about 10 feet. Altogether there were over a hundred good-sized pits in FIG. 13. HIDDEN GLACIEB IN 1905, SHOWING FOSSE BETWBBN THB GLACIEH AND THB WOBDB Kettle area. an area of about a square mile, and the region pitted by these kettles extended fully a mile from the visible glacier front. The presence of the kettles in the outwash gravel plain was interpreted by Gilbert, and by ourselves, as proof of the presence of ice beneath the gravels. There is abundant evidence to support this conclusion. In the first place, the great amount of clear, ice-cold water emerging from the kettles could be accounted for only by the melting out of buried ice. Numerous small streams issued from the kettles, their clear water contrasting strikingly with the muddy water of the nearby glacial streams. That the kettles were developing by irregular subsidence of the surface of the outwash gravel plain during OUT visit in 1905 was clearly proved by the abundant evidence of faulting and slumping of the gravels and the sliding down of the kettle walls. Furthermore, on the bottom of some of the kettles were deposits of mud recently made, and faulted by subsidence since their accumulation. The very existence of the kettles on the gravel plain is indication of the recency of their origin, for the surface of the plain was traversed by an intricate series of abandoned channelways, some of the largest with banks 80 or » Gilbert, G. K., Glaciers and Glacktion, Harriman Alaska Expedition, Vol. 8, 1904, p. 64.